The Reaction of Botulinum Toxin Type A with Ketene1

Chemical Corps, Fort Detrick]. The Reaction of Botulinum Toxin Type A with Ketene1. By Edward J. Schantz and. Leonard Spero. Received August 2, 1956...
0 downloads 0 Views 399KB Size
April !5, 1957

REACTION O F BOTULINUMTOXIN TYPEA

WITIT

KETENE

1623

[CONTRIBUTION FROM THE CHEMICAL CORPS, FORTDETRICK]

The Reaction of Botulinum Toxin Type A with Ketene] BY EDWARD J. SCHANTZ AND LEONARD SPERO RECEIVEDAUGUST2, 1956 When botulinum toxin, type A, was exposed t o ketene a rapid inactivation took place which followed first-order kinetics. The inactivation was accompanied by a small but continuous acetylation of free amino groups. Some 0-acetylation was evident but it was irregular and was not directly correlated with the decrease in toxicity. Toxin, completely inactivated by ketene, was ineffective in immunizing mice against the toxin.

The toxin produced by Clostridium botulinum, served here were best obtained with small volumes type A, has been isolated as a crystalline protein of solution in which the surface could be effectively and found to be homogeneous in electrophoretic renewed. The decrease in toxicity and free amino nitrogen and ultracentrifugal a n a l y s i ~ . ~A. ~ study of the composition of the toxin4 showed i t to be a simple brought about by acetylation with ketene is shown protein made up of nineteen L-amino acids linked in Table 1. It is evident that relatively small through normal peptide bonds with no evidence of changes in amino nitrogen occurred as the toxin was a prosthetic group. Presumably, therefore, the inactivated; thus a t 2.5 minutes 43% of the actoxicity must be due to the structural configuration tivity was lost when 5y0of the amino groups had reacted, and a t 5 minutes when 98% of the acwithin the molecule. The reaction of ketene with many proteins has tivity had disappeared, 19% of the amino groups been studied by a number of investigator^,^ and had reacted. A plot of the logarithm of the free information has been obtained on the relation of amino nitrogen against time yielded a straight line certain reactive groups within protein molecules to demonstrating a continuous and constant acetylabiological activity. Most work has centered on the tion throughout the reaction ; in addition, plotting reaction of ketene with free amino groups and with the fall in toxicity against the fall in free amino phenolic hydroxyl groups ; the reaction with sulf- nitrogen demonstrated a direct correlation between hydryl groups has been followed to a lesser extent. the two effects. The rate of decrease of toxicity This report presents evidence that ketene inac- with time of exposure would appear to be much tivates botulinum toxin by a reaction with a slower than in the experiment shown in Fig. 1. The difference is due to the much greater volume relatively small number of free amino groups. of toxin reacting in this experiment (48-fold) Results with equal rates of ketene supply. The greater Exposure of the toxin to ketene brought about volumes were necessitated in order to have suffia rapid fall in toxicity. The results of a typical cient material for the amino nitrogen determinaexperiment are shown in Fig. 1. A plot of the tions. logarithm of the toxicity against the time of ketene treatment pointed out a straight line relationship for almost a three log or 99.8yo loss in toxicity. The reaction therefore appears to be pseudounimolecular. The change in rate of inactivation during the last part of the exposure period may be attributed to a drop in pH which occurred a t the same time, since the rate of acetylation decreases as the pH is lowered.6 In these experiments i t was necessary to introduce the ketene above the surface of the liquid, instead of bubbling it through the solution as is usually done, in order to prevent surface denaturation of the toxin. Bubbling air or even an inert gas such as nitrogen through a solution of the toxin caused a rapid inactivation accompanied by the appearance of a large amount of insoluble protein.' The first-order kinetics obOO 20 40 60 (1) Presented, in p a r t , before the Division of Biological Chemistry, 128th Meeting, American Chemical Society, Minneapolis, Minn., September, 1955. (2) C. Lamanna, 0. E. McElroy and H. W. Eklund, Science, 103, 613 (1946). (3) A. Abrams, G. Kegeles and G . A. Hottle, J . B i d . Chem., 164, 63 (1946). (4) H. J. Buehler, E . J. Schantz and C. Lamanna, ibid., 169, 295 (1947). ( 5 ) H. S. Olcott and H. Fraenkel-Conrat, Chem. Reus., 41, 151 (1947); R. M. Herriott, Adwonces in Protein Chem., 3, 169 (1947). ( 6 ) R . M. Herriott and J. H. Northrop, J . Gen. Physiol., 18, 35 (1954). (7) I.. Spero and E. J. Schantz, unpublished.

EXPOSURE TIME IN SECONDS.

Fig. 1.-The effect of ketene on the toxicity of botulinum toxin, type A. The gas was introduced over the surface of 0.25-ml. samples of toxin in 2 cm. diameter vials.

The extent of 0-acetylation brought about by ketene inactivation of the toxin is shown in Table 11. The data indicated no reaction during the disappearance of one-third of the toxicity but between a disappearance of one-third and two-thirds some 0-acetylation became evident. The tryptophan equivalents in micrograms per mg. of toxin

EDWARD J. SCHANTZ AND LEONARD SPERO

1624

TABLE I THE DECREASEI N TOXICITY AND AMINO NITROGENOF BOTULISUMTOXIN TREATED WITH KETENE^ 13xposure time, min.

Toxicity, LDso/ml.

Free amino N. h y/mg. toxin

44 x 106 8.33 40 X 106 8.48 2.5 25 x 108 7.91 ;i. 0 12 x 106 7.44 10.0 1.05 X 106 6.75 In this experiment, 12-1111. samples were exposed t o ketene gas. I, A subsequent determination of the true value of free amino nitrogen (11. A . Rutherford, M. Harris and A . I,. Smith, J . Research AVuil. Bur. Standards, 19, 167 and the of lysine, gave a figure (1937)),that is CU-XHZ of 7.78 ylmg. of toxin, so that the values in the table are a little high. The error is due to the partial reaction of the guanidino group of arginine, but it should have been constant for all the determinations and, therefore, should not :itTect the interpretatiori of the data.

Vol. 7 ! )

aration did not flocculate with antitoxin and was ineffective as a toxoid. It provided no protection against as small a challenge dose as 10 LD,,. A formalinized toxoid of the same concentration protected against 4 X lo4LD6,,.8

(I

1.o

TABLE I11 THE EFFECTOF KETENEO N THE COMBINING POWER OF BOTULINUM TOXIN WITH ANTITOXIN Sample of toxin

Toxicity, LDdml.

Control 11 x 1 0 6 Ketene treated 4 . 0 x 106 Coiitrol 11 x 106 1.1 x 106 Ketene treated " A diflerent hntrli of antitoxin.

Flocculation Lf/ml. Kf,mill.

27 26-33 21" 21"

1'4 12 10 18

Discussion

The groups in proteins with which ketene usually reacts significantly are the free amino, phenolic relatively hydroxyl and sulfhydryl g r o ~ p s . The ~ minor loss in toxicity of botulinum toxin treated with p-chloromercuribenzoic acid serves to eliminate sulfhydryl groups as a primary reacting group in the inactivation induced by ketene. StaudinTABLE I1 ger'" states that primary amines react faster than THE0-ACETYLATION AND ACCOMPANYING Loss IN TOXICITYphenols and this observation has been generally found to apply to proteins5; furthermore, in those OF BOTULINJMTOXIS'BROUGIIT ,\BOUT BY KETENE' proteins in which phenolic hydroxyl groups were Exposure time, Toxicity, Tryptophan equiv., r / m g . toxin min. I.Ddml. necessary for activity there was a considerable 43.5 0 60 X lo6 time lag during exposure to ketene before inactiva44.6 0.5 >50 x 106 tion became apparent. 44.5 1 .o 40 x 1 0 6 The extremely rapid inactivation of botulinum 1.5 20 x 108 37.0 toxin by ketene suggests, therefore, a primary re37.7 2.0 13 X lo6 action with amino groups; the amino nitrogen and 3 0 < 2 . 5 X 106 37.4 phenol determinations support this interpretation. In this experiment, 4-1111. samples were exposed t o ke- The reduction in free aniino groups was continuous tene gas. throughout the inactivation, whereas there was no An independent measure of whether the sulfhy- measurable 0-acetylation until a considerable dedryl groups were involved in the ketene inactivation toxification had taken place. It is hypothesized of the toxin was made by observing the effect of that the essential groups are the most reactive to p-chloromercuribenzoic acid upon the toxicity of ketene, that they are few in number and that either botulinum toxin. In eight experiments in four the a-amino groups or a fern of the lysine €-amino separate runs, an average reduction in toxicity of groups are involved. One factor, however, prompts a cautious use of 32% was found in the presence of a t least 100 moles of reagent per mole of toxin. This effect was sig- the analytical data. If a very small number of nificant statistically but the magnitude was far less phenolic hydroxyl groups are essential in each than that observed in the inactivation with ketene. molecule, 0-acetylation of these groups might go The inactivation with p-chloromercuribenzoic acid undetected in the chemical determination. The may not have been specific since i t was only par- accuracy of the analytical method restricts the tially prevented by cysteine. Furthermore, an- number of undetectable groups to a maximum of other SH reagent, iodosobenzoic acid, had no effect two in a toxic sub-unit of 70,000 molecular weight" containing about fifty tyrosine residues. whatever on the toxicity. (8) G. A. Hottle, C. Xigg and J. A. Lichty, J. I m m u r m l . , 56, 255 The combining power of ketene-treated toxin with specific antiserum is presented in Table 111. (1947). I t should be observed a t this point t h a t the inactivation reaction A loss of two-thirds of the toxicity caused no change of (9) ketene may involve some other chemical function in t h e protein. in flocculation with the antitoxin and even a loss of T h e reactions followed in this work are certainly t h e primary ones in 9970 of the toxicity did not decrease the number of most cases b u t t h e possibility always exists t h a t there is present in the toxin a n essential grouping, usually unreactive or slowly reactive t o flocculating units although it did retard the rate of ketene in most proteins but highly reactive in this case. flocculation. This suggests that ketene might be (10) H. Staudinger, "Die Ketene," F. Enke, Stuttgart, 1912, p. 34. useful as a toxoid-producing reagent, but i t is (11) T h e toxin a s normally isolated has a molecular weight of apnecessary for safety purposes that an immunizing proximately one million, and it has been calculated t h a t t h e molecule contains 872 tyrosine residues (cf. ref. 4). Recent work (J. Wagman antigen be non-toxic; the 9970 inactivated mateand J. B. Bateman, Arch. Biochem. Biophrs., 45, 375 (1953): J. rial still had 1.1 X lo5 LDm per ml. Accordingly, Wagman, ibid., 50, 104 (1954)) indicates t h a t t h e toxic sub-unit of this cui alniost completely inactivated sample was pre- molecule has a molecular weight of about 70,000, which would contain pared by repeated ketene treatment. This prep- aboiit fifty tyrosine residues. dropped from about 44 to 37 and remained there throughout the experiment. The results of several other experiments showed considerable variation in the amount and time of 0-acetylation, but as in the experiment cited here there was no evidence of a direct correlation with t h e decrease in toxicity.

,

0

April 5, 1957

REACTION O F

BOTULINUM TYPEA WITH NITROUSACID

Experimental Materials .-The preparations of botulinum toxin, type A, used in this work were isolated by either the method of Abrams, et aZ.,3 or by a method based on differential ultracentrifugation.' Only crystalline preparations homogeneous in the ultracentrifuge at pH 3.8, with a specific activity, when isolated, of at least 200 X lo6 L D S Oper mg. N, were used. The toxin was stored in acetate buffer a t 5' and adjusted immediately before use with an appropriate phosphate buffer. The ketene was produced by the pyrolysis of acetone in an apparatus similar t o that designed by Herriott.l* The rate of production was measured by the neutralization of standard alkali under conditions of the actual acetylations. Toxicity.-The number of LD60 per ml. was determined by injecting 16-20 gram mice intraperitoneally with 0.5 ml. of toxin diluted with a sterile 1% disodium phosphate solution containing 0.270 gelatin and adjusted to pH 6.8. Six mice were used in each group and deaths were recorded for 96 hr. The per cent. kill was plotted against the dose on probit-log dose paper with the probit for 100% killed taken as the probit for five out of six killed plus probit unit and the probit of 0% killed taken as the probit of one out of six killed minus 1 / ~ probit unit.i3 The best straight line was fitted by inspection and the dose corresponding to probit 5 was read off the graph. Amino Nitrogen.-The extent of N-acetylation was ascertained by the nitrous acid amino nitrogen method of Van Slyke." The reaction was carried out a t 30-32" for 20 minutes in a reaction mixture containing 5 ml. of sample, 2 ml. of glacial acetic acid-saturated sodium acetate (1 :1) and 2 ml. of sodium nitrite (saturated a t 5 " ) . The use of a buffered reaction, first suggested by Rutherford, et aZ.,I6 serves t o decrease the volume of gas evolved and the magnitude of the blank analysis. Phenolic Hydroxyl.-The extent of tyrosine O-acetylation was determined by Herriott's modification of the Folin phenol method a t PH 8. The solutions were read a t 750 mp in a Beckman model DU spectrophotometer. The standard curve was made with tryptophan. Nitrogen.-Nitrogen in protein samples was determined by micro-Kjeldahl using a mercuric oxide catalyst and a 4hr digestion period. Flocculation.-One Lf unit represents the amount of toxin which gives the most rapid flocculation with one standard unit of antitoxin." The Kf is the time interval for the earliest flocculation. The determination was carried out a t 40" in a water-bath with 1 ml. of diluted toxin and volumes of antitoxin (100 units per ml.) varying from 0.10 to 0.41 ml. using a dilution factor of 1.25.

.

(12) R. M.Herriott, J . Gcn. Physiol., 18, 69 (1934). (13) E.S. Weiss, A m . J . Public Hcolth, 38, 22 (1948). (14) D.D. Van Slyke, J . Biol. Chcm., 83, 425 (1929). (15) Cf.footnote b, Table I.

1625

The Rate of Inactivation of Toxin by Ketene .-Samples of 0.25 ml. of toxin dissolved in 0.5 M potassium phosphate buffer a t pH 6.6 were placed in a 2 cm. diameter vial. The outlet tupe of the ketene generator was placed about l/*,cm. above the surface of the liquid and the sample was swirled slowly. After exposure for a predetermined period, the reaction was stopped by adding 4.75 ml. of gelatin-phosphate diluent. Under these conditions, the absorption rate of ketene into standard alkali was found t o be 0.36 mmole per minute. The Reduction in Amino Nitrogen Due to Ketene Treatment of the Toxin.-Twelve-ml. samples of toxin dissolved in 0.5 M potassium phosphate buffer a t pH 6.4 were placed in 30-ml. beakers and exposed to ketene. The solutions were stirred slowly with a magnetic stirrer. Samples were withdrawn for toxicity tests immediately after exposure and amino nitrogen was determined in duplicate analyses the same day. The Reduction in Phenolic Hydroxyl Due to Ketene Treatment of the Toxin.-Four-ml. samples of toxin dissolved in 0.5 M phosphate buffer a t pH 6.5 were placed in 30-ml. beakers and exposed to ketene. The samples were swirled during exposure. Samples were withdrawn for toxicity tests and phenol determinations immediately after exposure. The Effect of Sulfhydryl Reagents on the Activity of the Toxin.-Samples of 0.1 ml. of toxin and varying small volumes (0.025 to 0.20 ml.) of dilute solutions (1 X 10-3 to 1 X 10+ M ) of the reagent to be tested, usually p-chloromercuribenzoic acid, were added to enough 0.1 M phosphate buffer a t pH 7.0 to make a final volume of 2.0 ml. The mixtures had a pH of 6.8-6.9. The reaction was allowed to proceed for 15 minutes a t room temperature when samples were withdrawn for toxicity determinations. The Antigenicity of Ketene Treated Botulinum Toxin.A 2.5-ml. sample of toxin dissolved in 0.5 M phosphate buffer a t PH 6.5 was exposed to ketene for two minutes. The pH was then adjusted back to pH 6.5 with 0.03 ml. of 10 N NaOH. Four exposures were made on the same sample with pH adjustment after each. The final solution had less than 40 LD60 per ml. compared to an original activity of 35 X lo8 L D S Oper ml. The solution was diluted o ml. of original unitto a concentration of 1 X lo8 L D ~ per age and an alum precipitated toxoid was prepared according to the method of Nigg, et aZ.17 The toxoid was tested by the constant toxoid method of Hottle, et aZ.,* in mice.

Acknowledgment.-The authors are indebted to hfr. W. I. Jones, Jr., for some of the chemical determinations and to Mr. J. T. Duff for carrying out the flocculation tests. (17) C. Nigg, G. A. Hottle, L. L. Coriell, A. S. Rosenwald and G. W. Beveridge, J . Immunol., 6 6 , 245 (1947).

FREDERICK, MARYLAND

(16) Globulin modified antitoxin, Lederle.

[CONTRIBUTIONFROM THE CHEMICAL CORPS, FORT DETRICK]

The Reaction of Botulinum Toxin Type A with Nitrous Acid' BY LEONARD SPERO AND EDWARD J. SCHANTZ RECEIVEDAUGUST2, 1956 Botulinum toxin, type A, was inactivated rapidly by treatment with nitrous acid in the presence of excess nitrite. This reaction, presumably deamination, followed first-order kinetics. It is shown that the deamination of alanine under these conditions is also first order and that this is to be expected from the rate equation for the reaction. These results provide additional evidence for the essential nature of the free amino groups in t h e toxicity of this protein.

The reaction of botulinum toxin, type A, with ketene was described in the preceding paper,2 and i t was proposed on the basis of the rate of reaction and analysis of reacting groups that the free amino (1) Presented, in part, before the Division of Biological Chemistry, 128th meeting, American Chemical Society, Minneapolis, Minn., September, 1955. (2) E. J. Scbantz and L. Spero, THIS JOURNAL, 79, 1623 (1957.)

groups, rather than the phenolic hydroxyl groups, were involved in the inactivation. I n order to obtain more definitive information on this point, the reaction of the toxin with nitrous acid has been studied. This reaction, when carried out in the presence of excess nitrite, has been proposeda as a means of distinguishing between these two reactive (3) J. E. Little and M. L. Caldwell, J . B i d . Chem.. 147,229 (1943).